231 52 91KB
English Pages 13 Year 2005
JOURNAL OF MATHEMATICAL PHYSICS 46, 013506 ~2005!
E 6 unification model building. III. Clebsch–Gordan coefficients in E 6 tensor products of the 27 with higher dimensional representations Gregory W. Andersona) Department of Physics and Astronomy, Northwestern University, 2145 Sheridan Road, Evanston, Illinois 60208
Toma´sˇ Blazˇekb) Department of Physics and Astronomy, University of Southampton, Highfield, Southampton SO17 1BJ, United Kingdom
~Received 9 February 2001; accepted 12 December 2001; published online 10 January 2005! E 6 is an attractive group for unification model building. However, the complexity of a rank 6 group makes it nontrivial to write down the structure of higher dimensional operators in an E 6 theory in terms of the states labeled by quantum numbers of the standard model gauge group. In this paper, we show the results of our computation of the Clebsch–Gordan coefficients for the products of the 27 with irreducible representations of higher dimensionality: 78, 351, 3518 , 351, and 3518 . Application of these results to E 6 model building involving higher dimensional operators is straightforward. © 2005 American Institute of Physics. @DOI: 10.1063/1.1448134#
I. INTRODUCTION
E 6 is the minimal simple gauge group which could accommodate one family of the observed fermions, and a family of Higgs states, into a single gauge multiplet.1 Therefore, unification models based on E 6 can provide relationships for the measured charged fermion masses and quark mixing angles: 13 unrelated independent parameters of the standard model of elementary particles, and at the same time a small set of E 6 symmetric operators may relate the charged fermion data both to the masses and mixings in the neutrino sector and to the parameters of the Higgs sector. In this respect, E 6 provides a framework for the most economic unified supersymmetric theories. As is well known the key feature among the observed masses of the three generations of fermions is the intergenerational hierarchy. Any unified model has to explain the origin of the hierarchy in terms of the dynamics of the underlying theory. In E 6 models, the hierarchy can follow from the pattern of the symmetry breaking as the rank 6 group is broken down to the standard model gauge group, possibly in a succession of steps. The hierarchy may be explicitly realized in terms of higher dimensional operators containing the light states after the superheavy degrees of freedom are integrated out of the effective theory generated below the E 6 breaking scale M 6 . 2 From the technical point of view, the construction of higher dimensional E 6 symmetric operators and their structure in terms of the standard model states is a nontrivial task. Assuming that the light states occupy the fundamental 27-dimensional irreducible representation ~irrep!, a complete knowledge of the tensor products of the 27 irrep with larger irreps is required. For instance, the E 6 symmetry allows for a higher dimensional operator containing the product of three 27’s and a 78, suppressed by some heavy scale M H >M 6 . If the 78 acquires a vacuum expectation value ~vev! v 6 'M 6 and all three 27’s contain light states, such an operator contributes to the generation of fermion mass matrices. In particular, for the first two families it may generate a!
Electronic mail: [email protected] On leave of absence from the Dept. of Theoretical Physics, Comenius Univ., Bratislava, Slovakia; electronic mail: [email protected]
b!
0022-2488/2005/46(1)/013506/13/$22.50
46, 013506-1
© 2005 American Institute of Physics
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-2
G. W. Anderson and T. Blazˇek
J. Chem. Phys. 46, 013506 (2005)
entries suppressed by v 6 /M H . Yet, the predictivity of E 6 can only be utilized if the exact form of the singlet in 27 ^ 27 ^ 27 ^ 78 in terms of the standard model states is known. If two of the 27’s are contracted antisymmetrically, one needs to know the Clebsch–Gordan decomposition of the 351 in the tensor product 27 ^ 78, then the decomposition of the 27 in the product 27 ^ 351, and, lastly, the decomposition of the singlet in the product 27 ^ 27. However, complete information on general tensor products of the exceptional group E 6 is difficult to obtain.3 As for particular computations, to our knowledge only the Clebsch–Gordan decompositions of 27 ^ 27, 27 ^ 27, and 78 ^ 78 are presently available in the literature.4 – 6. ~We note in passing that a separate paper7 obtains a subset of the results needed for the operator 27378 by studying a branching chain of E 6 . The results presented in this paper are relevant for the case when the vev is acquired by a zero weight state of the 78.! In this paper, we continue in our earlier work5,6 and provide basic group-theoretical tools for a construction of higher-dimensional E 6 symmetric operators. In particular, we present the results of our computation of the Clebsch–Gordan coefficients ~CGCs! for the tensor products involving 27-, 78-, and 351-dimensional representations, the lowest dimensional irreps in E 6 . Section II contains some necessary mathematical background for our study, mostly concerned with lowering in the weight system of these irreps. Our main results can be found in Sec. III, where we also comment on the construction and properties of the weight systems of the resulting irreps. Section IV contains the summary, while the appendix provides details on the lowering relations in the presence of degenerate weights.
II. MATHEMATICAL PRELIMINARIES
In this work we consider tensor products of the fundamental 27-dimensional irrep with higher dimensional 78- and 351-dimensional representations of E 6 . In particular, we compute the Clebsch–Gordan coefficients for the products 27 ^ 7851728 % 351 % 27,
~1a!
27 ^ 35157371 % 1728 % 351 % 27,
~1b!
27 ^ 3518 57722 % 1728 % 27,
~1c!
27 ^ 35155824 % 2925 % 650 % 78,
~1d!
27 ^ 351853003 % 5824 % 650.
~1e!
Before we discuss the construction of the weight systems of the irreps on the right-hand side of these relations let us start first with the rules for the construction of the irreps on the left. The key ingredient of our procedure is the lowering operation which is used to construct a complete weight system by successive application of generators E 2 a 1 ,...,E 2 a 6 . These are the generators which lie outside of the diagonal Cartan subalgebra and correspond to the six simple roots of E 6 . ~Our choice of generators is described in more detail in Ref. 6, and basically follows the standard conventions of Ref. 8.! The six generators act as ladder operators—at each level the weight of the new state is obtained from the weight at the previous level by subtracting ~in the weight space! the respective simple root: E 2 a i u w & 5N 2 a i ,w u w2 a i & .
~2!
For the weight systems of the 27 and 78 constants N 2 a i ,w satisfy @see Eq. ~12! in Ref. 6# u N 2 a i ,(w) j u 2 5 ^ a i u w & 1 u ^ ~ w ! j u ~ w ! i & u 2 u N 2 a i ,w1 a i u 2 .
~3!
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-3
E 6 unification model building. III
J. Chem. Phys. 46, 013506 (2005)
It is understood that the new state u w2 a i & does not exist if N 2 a i ,w 50 or the right-hand side of ~3! turns out to be negative. The subscript on the weight ~w! is only relevant for the six degenerate zero weight states of the 78 and is to be ignored for nondegenerate weights. In fact, the second term on the right-hand side of ~3! never contributes when one constructs the weight system of the 27 as there are no higher multiplets than doublets for any SU(2) subgroup. We remark that throughout this work, and consistent with our previous studies,5,6 the lowering phase convention which always fixes constants N to be real and non-negative N 2 a i ,w >0
~4!
is adopted for any simple root a i and any weight system. Then for the zero weight states of the 78 the inner product in ~3! can be expressed as
^ ~ w ! i u ~ w ! j & 5 u A i j u /2,
~5!
where A i j [ ^ a i u a j & are the elements of the Cartan matrix of E 6 . 8,6 This result follows from the decomposition of the 78 weight states into the states of the fundamental representations in the tensor product 27 ^ 27. 5 In the appendix, we derive a generalized relation for N 2 a i ,w for a weight system with multiple degenerate weights at different levels. We now discuss how to apply general formula ~A8! to the weight systems of the 351-dimensional representations which appear on the left in Eqs. 1~b!–1~e!. These irreps, although already rather large, are still special because for each weight subspace a basis can be defined such that the application of a lowering ladder operator results in a single basis state, as indicated in Eq. ~2!. ~For larger irreps, there are lowerings which lead to a linear combination of the basis states regardless of the basis definition.! Moreover, if weight ~w! is degenerate and a state with weight (w2 a ) exists, then the (w1 a ) weight state is either nondegenerate or does not exist at all. Thus for the weight system of the 3518 or 351, Eq. ~A8! reduces to a simple form u N 2 a ,(w) a →(w2 a ) A u 2 5 ^ a u w & 1 u ^ ~ w ! a u ~ w ! c & u 2 u N 2 a ,(w1 a )→(w) c u 2 ,
~6!
a
where, formally, the summation over c is assumed in the last term, but no more than one state actually contributes. Concrete applications of this formula are provided at the end of the section. Compared to Eq. ~3! both weights ~w! and (w2 a ) can now be degenerate. We find, however, that in the 3518 or 351 the (w1 a ) weight state does not exist if (w2 a ) is a degenerate weight. Hence if both ~w! and (w2 a ) are degenerate, A a can be set to a by definition and ~6! can be simplified even further: u N 2 a ,(w) a →(w2 a ) a u 2 5 ^ a u w & ,
for both ~ w ! and ~ w2 a ! degenerate.
~7!
This shows that the definition of the basis states ~and their subscript labeling! in the degenerate subspaces of the 3518 or 351 can be induced from the basis states at the previous level. However, once ~w! is found degenerate, how do we know if (w2 a ) is going to be degenerate and what the dimensionality of this subspace is going to be? Similar to the case of the 78 weight system6 the decomposition of the 351-dimensional irreps into the states of the fundamental representations can be recalled. The product 27 ^ 2753518 % 351 % 27 is conjugated to the product studied in Ref. 4. We refer to this work to claim that all degenerate weight subspaces in the 3518 ~or 351! are of the same dimensionality and that the degenerate weights follow the weight system of the 27. In the end it thus turns out that complete bases in the degenerate weight subspaces of the 3518 can be obtained starting from the four ~100000! weight states at level 8 of the 3518 :9 ¯ 21 ¯ 01 ¯ ! & /&, u ~ 100000! 3 & 5E 2 a 3 u ~ 11
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-4
G. W. Anderson and T. Blazˇek
J. Chem. Phys. 46, 013506 (2005)
¯ 21 ¯ 0 ! & /&, u ~ 100000! 4 & 5E 2 a 4 u ~ 101 ~8! ¯ 20! & /&, u ~ 100000! 5 & 5E 2 a 5 u ~ 1001 ¯ 002! & /&. u ~ 100000! 6 & 5E 2 a 6 u ~ 101 With lowering convention ~4! the remaining 26 degenerate weight subspaces at lower levels can be specified as u (w) a & 5E 2 a A ¯E 2 a B u (100000) a & , a53,4,5,6, where E 2 a A ¯E 2 a B is the lowering path leading to state u (w) & in the 27. For the 351 the only difference is that the degenerate weight subspaces are five-dimensional and the relations analogous to ~8! also include ¯ 000! & /& u ~ 100000! 2 & 5E 2 a 2 u ~ 021
~9!
when computing the ~100000! states at level 7 of this irrep. We note that with this notation the inner product in any degenerate weight subspace of both the 3518 and 351 satisfies
^ ~ w ! a u ~ w ! b & 5 u A ab u /2,
~10!
~where a,b53,4,5,6 for the 3518 , and a,b52,3,4,5,6 for the 351!, in a close similarity to the degenerate zero weight subspace of the 78, Eq. ~5!. As an example of the application of formula ~6! consider all possible lowerings of the state u F 5 & in the 3518 where, for brevity, F stands for the ~100000! weight. Three different states at the ¯ 10000) 5 & , E 2 a u F 5 & 5N 4 u 1012 ¯ 10& , and next level can be obtained: E 2 a 1 u F 5 & 5N 1 u (1 4 ¯ 0 & . Based on ~6! the constants are E 2 a u F 5 & 5N 5 u 10012 5
u N 1 u 2 511051, u N 4 u 2 501
SD 1 2
2
1 ~ & !25 , 2
u N 5 u 2 5011 2 ~ & ! 2 52,
as we have already shown in the second and third equations of ~8! that N 2 a 4 ,(F1 a 4 )→(F) 4 5N 2 a 5 ,(F1 a 5 )→(F) 5 5&. Implicitly, we also used the fact that ~8! represents the only way the ~100000! weight states can be obtained from the states at the previous level. Note that ~7! could be used to compute N 1 since ¯ 10000) are degenerate weights. both ~100000! and (1 Finally, we remark that the properties of the 3518 and 351 are easily derived from the properties of the 3518 and 351 after the Dynkin coordinates @and any other indices in Dynkin formalism, like e.g., the labeling of states in Eq. ~8!# 1 and 2 are exchanged with 5 and 4, respectively.
III. CONSTRUCTION OF CLEBSCH–GORDAN COEFFICIENTS
Tensor products in Eq. ~1! can be expressed in terms of the highest weights as ~ 100000! ^ ~ 000001! 5 ~ 100001! % ~ 000100! % ~ 100000! ,
~11a!
~ 100000! ^ ~ 000100! 5 ~ 100100! % ~ 000011! % ~ 010000! % ~ 000010! ,
~11b!
~ 100000! ^ ~ 000020! 5 ~ 100020! % ~ 000011! % ~ 000010! ,
~11c!
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-5
E 6 unification model building. III
J. Chem. Phys. 46, 013506 (2005)
TABLE I. Bases in the dominant weight subspaces of the 1728-dimensional ~100001! irrep. Weight state
Lowering path
u 0001006 & u 0001003 & u 0001002 & u 0001001 &
6321 3621 2361 1236
u 1000001 & u 1000002 & u 1000003 & u 1000004 & u 1000005 &
12364534236 21364534236 25364436321 24534636321 23643254361
Weight state u 1000006 & u 1000007 & u 1000008 & u 1000009 & u 10000010& u 10000011& u 10000012& u 10000013& u 10000014& u 10000015& u 10000016&
Lowering path 65324436321 64534236321 63214534236 63243654321 53624436321 54321634236 32164534236 32643654321 32643254361 45346236321 45321634236
~ 100000! ^ ~ 010000! 5 ~ 110000! % ~ 001000! % ~ 100010! % ~ 000001! ,
~11d!
~ 100000! ^ ~ 200000! 5 ~ 300000! % ~ 110000! % ~ 100010! .
~11e!
For each product we start with the construction of the weight system of the first irrep on the right-hand side. The highest weight state of this irrep is nondegenerate and can always be ex-
TABLE II. Bases in the dominant weight subspaces of the ~100100! irrep, the 7371. u 000010n & states are marked u ¯F n & for brevity. Weight state
Lowering paths to ~000010! weight states
Lowering path
u 0000114 & u 0000113 & u 0000112 & u 0000111 &
4321 3421 2341 1234
u 2000006 & u 2000003 & u 2000004 & u 2000005 & u 2000002 &
6345234 3645234 4365234 5436234 2364534
u 0100001 & u 0100002 & u 0100003 & u 0100004 & u 0100005 & u 0100006 & u 0100007 & u 0100008 & u 0100009 & u 01000010& u 01000011& u 01000012& u 01000013& u 01000014& u 01000015&
23645341 32645341 31645234 35644321 36435421 61345234 65434321 64534321 63245341 41365234 43265341 43546321 51436234 54326341 12364534
u¯ F 1& u¯ F 2& u¯ F 3& u¯ F 4& u¯ F 5& u¯ F 6& u¯ F 7& u¯ F 8& u¯ F 9& u¯ F 10& u¯ F 11& u¯ F 12& u¯ F 13& u¯ F 14& u¯ F 15& u¯ F 16& u¯ F 17& u¯ F 18& u¯ F 19& u¯ F 20& u¯ F 21& u¯ F 22&
514362236434321 563214436234321 536214436234321 523614436234321 145362236434321 146234536234321 143621236345234 143622336435421 162363245434321 162332143645234 162332435644321 134562236434321 134632364523421 134621236345234 134622336435421 121364236345234 122334546634321 123456321436234 624134536234321 621363245434321 621332143645234 621332435644321
u¯ F 23& u¯ F 24& u¯ F 25& u¯ F 26& u¯ F 27& u¯ F 28& u¯ F 29& u¯ F 30& u¯ F 31& u¯ F 32& u¯ F 33& u¯ F 34& u¯ F 35& u¯ F 36& u¯ F 37& u¯ F 38& u¯ F 39& u¯ F 40& u¯ F 41& u¯ F 42& u¯ F 43& u¯ F 44&
645342136234321 643452136234321 643621345234321 636231245434321 633221143645234 632145364234321 344523126634321 343221166345234 345234126634321 345216321436234 342231166345234 342345126634321 342312632645341 342163423546321 245134326634321 241345326634321 241332166345234 213245346634321 213243632645341 213456321436234 453423126634321 432163423546321
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-6
G. W. Anderson and T. Blazˇek
J. Chem. Phys. 46, 013506 (2005)
TABLE III. Bases in the dominant weight subspaces of the ~100020! irrep, the 7722. ~100100! weight is left out as trivial. u 000010n & states are marked u ¯ F n & for brevity. Weight state
Lowering paths to ~000010! weight states
Lowering path
u 0000115 & u 0000114 & u 0000113 & u 0000112 & u 0000111 &
54321 45321 34521 23451 12345
u 2000006 & u 2000003 & u 2000004 & u 2000005 &
63452345 36452345 43652345 54362345
u 0100001 & u 0100002 & u 0100003 & u 0100004 & u 0100005 & u 0100006 & u 0100007 & u 0100008 & u 0100009 & u 01000010& u 01000011& u 01000012& u 01000013& u 01000014&
263453451 163452345 136452345 143652345 154362345 653454321 645345321 633454521 326453451 356454321 344655321 455364321 453263451 543263451
u¯ F 1& u¯ F 2& u¯ F 3& u¯ F 4& u¯ F 5& u¯ F 6& u¯ F 7& u¯ F 8& u¯ F 9& u¯ F 10& u¯ F 11& u¯ F 12& u¯ F 13& u¯ F 14& u¯ F 15& u¯ F 16& u¯ F 17& u¯ F 18& u¯ F 19& u¯ F 20&
5632144362345321 5362144362345321 5236144362345321 5123644362345321 5432163452364321 6421345362345321 6412345362345321 6453421362345321 6434521362345321 6212363343454521 6213632343454521 6213321436452345 6213324356454321 6136232343454521 6134522363454321 6134213623452345 6334542361234521 6321453623454321 3164362123452345 3164522363454321
u¯ F 21& u¯ F 22& u¯ F 23& u¯ F 24& u¯ F 25& u¯ F 26& u¯ F 27& u¯ F 28& u¯ F 29& u¯ F 30& u¯ F 31& u¯ F 32& u¯ F 33& u¯ F 34& u¯ F 35& u¯ F 36& u¯ F 37& u¯ F 38& u¯ F 39& u¯ F 40&
3164213623452345 3164223645345321 3445231266345321 3452341266345321 3452163452364321 3452163421362345 3422636345123451 3423451266345321 3423126633454521 2451343266345321 2411363623452345 2413453266345321 2413322663453451 2132453466345321 2136453421362345 1453622363454321 1423453266345321 1236453421362345 4534231266345321 4532163452364321
pressed as a trivial combination of the highest weight states of the two irreps on the left-hand side, with the CGC being equal to 11. In the absence of a simple method to determine the bases in the degenerate weight subspaces which follow at lower levels for each of these irreps, we compute directly the complete weight system in each case. However, note that simple lowering ~2! does not necessarily hold for weights with multiple degeneracies, as discussed in the appendix. States at lower levels are then computed by successive lowerings applied to the states of the 27, and 78 in case ~a! or one of the 351-dimensional irreps in cases ~b!–~e!. These lowerings were described in detail in Sec. II. The computed state is accepted and kept as a new basis state if it cannot be expressed as a linear combination of the previously obtained basis states with the same weight. It is not necessary to show the Clebsch–Gordan coefficients for every linearly independent state, since there are many states with the same CGCs. Instead, we present the results just for the dominant weight states. Dominant weights are weights with all Dynkin coordinates non-negative. The CGCs for the remaining states can then be determined using the charge conjugation operators,10,11 or in a straightforward way by direct lowering. In Tables I–V we present lowering paths for the dominant weight states of the 1728, 7371, 7722, 5824, and 3003 irreps. In our abbreviated notation, lowering path, let’s say, 3421 stands for E 2 a 3 E 2 a 4 E 2 a 2 E 2 a 1 applied ~from the right! to the highest weight state. The lowering paths in Tables I–V actually specify our choice of bases for particular dominant weight subspaces. Explicit Clebsch–Gordan decomposition of the dominant weight states is important because, typically, the multiplicity of degeneracy ~i.e., the dimensionality of the weight subspace! changes compared to the degeneracy at the previous level. Clearly, that is why these states cannot be obtained by generalized charge conjugation from the
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-7
E 6 unification model building. III
J. Chem. Phys. 46, 013506 (2005)
TABLE IV. Bases in the dominant weight subspaces of the ~110000! irrep, the 5824. Weight state
Lowering path
u 0010002 & u 0010001 &
21 12
u 1000101 & u 1000102 & u 1000103 & u 1000104 & u 1000105 & u 1000106 & u 1000107 & u 1000108 &
1236432 2136432 2364321 6433221 6321432 4321632 3216432 3264321
u 0000011 & u 0000012 & u 0000013 & u 0000014 & u 0000015 & u 0000016 & u 0000017 & u 0000018 & u 0000019 & u 00000110& u 00000111& u 00000112& u 00000113& u 00000114& u 00000115& u 00000116& u 00000117& u 00000118& u 00000119& u 00000120& u 00000121& u 00000122& u 00000123& u 00000124&
653214433221 645342133221 536214433221 532144321632 523614433221 523144321632 512364433221 512344321632 544362133221 415321236432 415322364321 425132136432 425132364321 433654232121 436542133221 364354232121 314521236432 314522364321 322436543121 321432654321 211342365432 213243654321 213645321432 123645321432
Lowering paths to zero weight states u 0 1& u 0 2& u 0 3& u 0 4& u 0 5& u 0 6& u 0 7& u 0 8& u 0 9& u 0 10& u 0 11& u 0 12& u 0 13& u 0 14& u 0 15& u 0 16& u 0 17& u 0 18& u 0 19& u 0 20& u 0 21& u 0 22& u 0 23& u 0 24& u 0 25& u 0 26& u 0 27& u 0 28& u 0 29& u 0 30& u 0 31& u 0 32&
65241364345236342133221 65241363344521322364321 65142364345236342133221 65142363344521322364321 65362312453436214433221 65321453623436214433221 65321443632314521236432 65321443632314522364321 62451364345236342133221 62451363344521322364321 62113344223366554433221 62136324436321554433221 62133221443366554433221 62133214432635544321632 64152364345236342133221 64152363344521322364321 64536421345236342133221 64536213344521322364321 64534532364312364232121 61236324436321554433221 61233221443366554433221 61233214432635544321632 63623124436321554433221 63322114432635544321632 52331266453436214433221 52361236453436214433221 52361234536436214433221 52361453623436214433221 52361443632314521236432 52361443632314522364321 51362362453436214433221 51362324536436214433221
u 0 33& u 0 34& u 0 35& u 0 36& u 0 37& u 0 38& u 0 39& u 0 40& u 0 41& u 0 42& u 0 43& u 0 44& u 0 45& u 0 46& u 0 47& u 0 48& u 0 49& u 0 50& u 0 51& u 0 52& u 0 53& u 0 54& u 0 55& u 0 56& u 0 57& u 0 58& u 0 59& u 0 60& u 0 61& u 0 62& u 0 63& u 0 64&
51453643212233664433221 51436213452233664433221 51436213422334566321432 53623124536436214433221 53621453623436214433221 53621443632314521236432 53621443632314522364321 54433221166345322364321 24512345632133664433221 24513245632133664433221 24513246321334566321432 24513213466345321236432 24513213466345322364321 23312435454634216633221 23245341166345321236432 23245341166345322364321 23611435422334566321432 23612344321635544321632 41534563212233664433221 41534632122334566321432 41536213452233664433221 41536213422334566321432 45334221166345321236432 45334221166345322364321 45345632364312364232121 45346323465121322364321 13456342122334566321432 13456213452233664433221 13456213422334566321432 13456223645363214433221 36231244332166554433221 36231244321635544321632
states at the previous levels. Moreover, it is important to check the completeness of a reducible dominant weight subspace. If it is impossible to complete its basis by lowering the states at the previous level, new weight systems open up and the remaining basis vectors are their highest weight states. This is what happens for every dominant weight in the tensor products 78 ^ 78,6 27 ^ 27,5 or 27 ^ 275 studied in the earlier work. However, this property of the dominant weights is no longer true for the products studied here. We now briefly discuss the dominant weights in each of the products in ~11!. ~a! (100000) ^ (000001)5(100001) % (000100) % (100000). At level 4 of the 1728-dimensional ~100001! irrep we find four states with weight ~000 100!. This weight space, however, is fivedimensional, and the computation of the state orthogonal to the previous four yields the highest weight state of the 351 irrep. ~See Table VI.! Similarly, at level 11 we find 16-fold degenerate weight ~100000!, while this reducible subspace unfolds to be 22-dimensional. Since there are five distinct states of the same weight in the 351, there is room for one extra state. Once computed as orthogonal to all the other 21 states it becomes the highest weight state of the fundamental 27-dimensional ~100000! irrep. Because the subsequent tables of CG coefficients become too unwieldy to print, we have deposited electronic versions with the electronic Physics Auxiliary
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-8
G. W. Anderson and T. Blazˇek
J. Chem. Phys. 46, 013506 (2005)
TABLE V. Bases in the dominant weight subspaces of the 3003-dimensional ~300000! irrep. ~110000! weight is left out as trivial. Weight state
Lowering path
u001000&
211
u 1000106 & u 1000103 & u 1000102 & u 1000101 &
64332211 36432211 23643211 12364321
u 0000011 & u 0000012 & u 0000013 & u 0000014 & u 0000015 & u 0000016 & u 0000017 & u 0000018 & u 0000019 & u 00000110&
6544333222111 5364433222111 5236443322111 5123644332211 4251364332211 4152364332211 4536433222111 3145236432211 3452364322111 2345123643211
Weight state
Lowering path
u 0000001 & u 0000002 & u 0000003 & u 0000004 & u 0000005 & u 0000006 & u 0000007 & u 0000008 & u 0000009 & u 00000010& u 00000011& u 00000012& u 00000013& u 00000014& u 00000015& u 00000016& u 00000017& u 00000018& u 00000019& u 00000020& u 00000021& u 00000022& u 00000023& u 00000024&
652413643453323643222111 651423643453323643222111 653623124536444333222111 653214436323145236432211 624513643453323643222111 621134342362365544332211 621332243643655443322111 641523643453323643222111 645364332345221236432111 612332243643655443322111 524536361236444333222111 524133214663452364332211 524133245434666333222111 513645236236444333222111 514362134523623644332211 536214436323145236432211 245133214663452364332211 245133245434666333222111 213216324364365544332211 415345234234666333222111 415362134523623644332211 453342211663452364332211 134562134523623644332211 345634234653221236432111
Publication Service.14 Note that in Table VII14 we keep the labeling of the five ~100000! states of the 351 consistent with the notation introduced in Eqs. ~8! and ~9!. ~b! (100000) ^ (000100)5(100100) % (000011) % (010000) % (000010). Lowering down to level 4 of the 7371-dimensional ~100100! irrep we obtain four distinct ~000011! weight states spanned over a five-dimensional reducible subspace. The last basis state in this subspace, orthogonal to the four from the 7371, becomes the highest weight state of the 1728. ~See Table VIII.! This is a conjugate irrep to the 1728 described in ~a!. The lowering paths to the dominant weights in its weight system can be obtained from Table I ~replacing 1 and 2 with 5 and 4, and vice versa. Proceeding to level 7 a fivefold degenerate ~200000! dominant weight is found: u 200000a & 5 u 100000& u 100000a & ,
~12!
a52,...,6
TABLE VI. CG coefficients for ~000100! dominant weight in (100000) ^ (000001). Each entry should be divided by the respective number in the last row to keep the states normalized to 1. (100001) u 0001006 &
¯ & u 000001& u 000101 ¯ 101& u 001001 ¯& u 001 ¯ 1000& u 011 ¯ 100) u 01 ¯ 0100& u¯1 10000& u 11 ¯ u 100000& u 1 00100&
1 1
&
u 0001003 &
1 1
&
(000100) u 0001002 &
1 1
&
u 0001001 &
u000100&
1 1
1 21 1 21 1
&
A5
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-9
E 6 unification model building. III
J. Chem. Phys. 46, 013506 (2005)
TABLE VIII. CG coefficients for ~000011! dominant weight in (100000) ^ (000100). Each entry should be divided by the respective number in the last row to keep the states normalized to 1. (100100) u 0000113 &
u 0000114 &
¯ 11& u 000100& u 0001 ¯ ¯ 10& u 001 101& u 0011 ¯ ¯ u 01 1000& u 011 011& ¯ 0011& u¯1 10000& u 11 u 100000& u¯1 00011&
1 1
(000011) u 0000112 &
1 1
&
1 1
&
&
u 0000111 &
u000011&
1 1
1 21 1 21 1
&
A5
which, obviously, does not leave any extra space for states outside of the 7371. This is consistent with no observation in ~a! of a dominant weight ~000020! in the weight system of the 1728, and also with the fact that there is no ~200000! irrep on the right-hand side of ~11b!. The charge conjugation operators can be used to show that a fivefold degenerate weight subspace with CGCs ¯ 0) equal to 1 is then present at odd levels of the 7371 from this level down until subspace (00002 emerges at level 39. Next, at level 8 fifteen linearly independent ~010000! weight states are present, while the weight subspace turns out to be 20-dimensional ~Table IX!.14 Not surprisingly there are four states which belong to the weight system of the 1728 @compare with ~a!#, and the remaining basis state, orthogonal to the previous 19, represents the highest weight state of the 351. Last, at level 15 we get the reducible ~000010! weight subspace, which is 66-dimensional. That makes room for the highest weight of the 27, since there are 44 basis states present in the 7371 together with 16 states of the 1728. An additional five states of the 351 should be expected based on Eqs. ~8! and ~9!. The CGCs for this subspace are presented in Tables X and XI.14 ~c! (100000) ^ (000020)5(100020) % (000011) % (000010). In the construction of the 7722dimensional ~100020! irrep one finds a dominant weight already at level 1, u 100100& 5 u 100000& u 000100& .
~13!
It occupies a one-dimensional subspace, which is consistent with the absence of the ~100100! irrep in product ~11c!. The first degenerate dominant weight is obtained at level 5. There, a sixdimensional ~000011! weight subspace contains five linearly independent states of the 7722. The basis in this reducible subspace is completed by the highest weight state of the 1728 ~Table XII!. Proceeding further, there is no room for the highest weight state of a new irrep when dominant weights ~200000! and ~010000! are encountered at levels 8 and 9, respectively. The ~200000! TABLE XII. CG coefficients for ~000011! dominant weight in (100000) ^ (000020). Each entry should be divided by the respective number in the last row to keep the states normalized to 1. (100020) u 0000115 &
¯ 1 & u 000020& u 00001 ¯ 11& u 000100& u 0001 ¯ 101& u 0011 ¯ 10& u 001 ¯ 1000& u 011 ¯ 011& u 01 ¯ 0011& u¯1 10000& u 11 ¯ u 100000& u 1 00011&
1 &
)
u 0000114 &
1 1
&
u 0000113 &
1 1
&
(000011) u 0000112 &
1 1
&
u 0000111 &
u000011&
1 1
2& 1 21 1 21 1
&
A7
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-10
G. W. Anderson and T. Blazˇek
J. Chem. Phys. 46, 013506 (2005)
TABLE XVI. CG coefficients for ~001000! dominant weight in (100000) ^ (010000). Each entry should be divided by the respective number in the last row to keep the states normalized to 1. (110000) u 0010001 &
¯ 1000& u 010000& u 01 ¯ ¯ 1000& u 1 10000& u 11 ¯ u 100000& u 1 01000&
1 1 &
(001000) u 0010002 &
u001000&
1 1
1 21 1
&
)
subspace is four-dimensional and its basis can be specified as in Eq. ~12!. ~The states are now numbered as a53,4,5,6.! The CGC decomposition of the ~010000! subspace can be found in Table XIII.14 Finally, at level 16 the last dominant weight in this product is unveiled. The reducible ~000010! weight subspace turns out to be 57-dimensional, with 40 basis states coming from the 7722 and 16 states from the 1728. The remaining state, orthogonal to them, becomes the highest weight state of the 27 ~see Tables XIV and XV!.14 ~d! (100000) ^ (010000)5(110000) % (001000) % (100010) % (000001). In this product, we find the dominant weight states encountered already in the decomposition of 78 ^ 78 and 27 ^ 27. At level 2 of the ~110000! weight system ~i.e., the 5824 irrep! we reach the three-dimensional ~001000! subspace, with two states in the 5824 and the third one being the highest weight state of the 2925, as shown in Table XVI. Then following the lowering paths in Table IV, Table II in Ref. 6, and Table I in Ref. 5 the dominant weights ~100010!, ~000001!, and ~000000! follow at levels 7, 12, and 23, respectively. The CGCs for these dominant weights can be found in Tables XVII–XXIII.14 The reducible ~000000! subspace is 135-dimensional and represents the most ~technically! involved computation in this study. Obviously, it cannot ~and does not! leave any room for the singlet since the two representations in the product are not conjugate to each other. ~e! (100000) ^ (200000)5(300000) % (110000) % (100010). The 3003-dimensional ~300000! irrep contains a dominant weight already at level 1: u 110000& 5 ~ u¯1 10000& u 200000& 1& u 100000& u 010000& )/).
~14!
The orthogonal combination u 110000& 5 ~ & u¯1 10000& u 200000& 2 u 100000& u 010000& )/)
~15!
forms the highest weight state of the 5824. Since then, the same dominant weights occur as in the weight system of the 5824 described under ~d!. There are, however, no 2925 and 78 irreps in this product ~see Tables XXIV–XXVI14!, just the highest weight state of the 650 completes the 13dimensional ~100010! subspace at level 8 ~Table XXVII14!. The reducible ~000000! weight subspace is 108-dimensional and its decomposition can be found in Tables XXVIII–XXXI.14
IV. SUMMARY
We have presented the Clebsch–Gordan decomposition of the E 6 tensor products of the fundamental 27 irrep with the 78- and 351-dimensional irreps. Analogous products involving the 27 instead of the 27 can now be obtained trivially by charge conjugation. It is straightforward to apply these results to the construction of higher dimension operators in E 6 model building.12
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-11
E 6 unification model building. III
J. Chem. Phys. 46, 013506 (2005)
APPENDIX THE PROBLEM OF DEGENERATE WEIGHTS
Rules ~2,3! are insufficient for representations with degenerate weights at successive levels. For degenerate weights we must first identify a particular basis. Label the degenerate basis states of weights (w1 a ), (w), and (w2 a ) as u ~ w1 a ! G & , u w c& ,
G51,...D w1 a ,
where where
c51,...D w ,
and u ~ w2 a ! C & ,
where
C51,...D w2 a .
D w stands for the degeneracy of (w). The basis states are in general nonorthogonal. In our notation, they are always normalized to unity: ^ w c u w c & 51. The identity operator in the degenerate subspace is I5G ab u w a &^ w b u , ~A1! G ab 5 ~ M 21 ! ab ,
where
M ab 5 ^ w a u w b & .
Although the basis is nonorthogonal, we can construct state vectors which are orthogonal to any state except the state we are interested in ˆ b & 5 u w a & G ab , uw ~A2!
^ wˆ a u 5G ab ^ w b u , which satisfy
^ w c u wˆ a & 5 ^ wˆ a u w c & 5 d ac .
~A3!
A general raising or lowering of a degenerate weight state can be written as E a i u w c & 5N a i ,w c →(w1 a i ) G u ~ w1 a i ! G & ,
~A4!
E 2 a i u w c & 5N 2 a i ,w c →(w2 a i ) C u ~ w2 a i ! C & ,
~A5!
where there is a possible sum over the states on the right-hand side @compare ~A5! with ~2!#. The lowering normalization constant can then be expressed only as a sum of matrix elements (w2 a ) † N 2 a i ,w a →(w2 a i ) A 5G AB i ^ (w2 a i ) B u E 2 a i u w a & . 13 Using E a 5E 2 a and the defining relation ~A5! we derive
* a ,w N 2 a ,w a →(w2 a ) A N 2
b → ~ w2 a ! B
^ ~ w2 a ! B u ~ w2 a ! A &
5 ^ w bu E aE 2au w a& 5 ^ w b u @ E a ,E 2 a # 1E 2 a E a u w a & w1 a 5 ^ w b u w a &^ a ,w & 1G GD ^ w b u E 2 a u ~ w1 a ! G &^ ~ w1 a ! D u E a u w a & w1 a * a , ~ w1 a ! 5 ^ w b u w a &^ a ,w & 1G GD N 2 a ,(w1 a ) G →w c ^ w b u w c & N 2
D →w d
^ w du w a& .
~A6!
Hence
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-12
G. W. Anderson and T. Blazˇek
J. Chem. Phys. 46, 013506 (2005)
21 * a ,w N 2 a ,w a →(w2 a ) A N 2 ~ G w2 a ! AB
b → ~ w2 a ! B
21 21 w1 a 21 * a , ~ w1 a ! N 2 a ,(w1 a ) G →w c N 2 5 ~ G w ! ab ~ G w ! da ~ G w ! bc ^ a ,w & 1G GD
D →w d
.
~A7!
For a5b we get 21 * a ,w N 2 a ,w a →(w2 a ) A N 2 ~ G w2 a ! AB
a → ~ w2 a ! B
w1 a 21 21 * a , ~ w1 a ! N 2 a ,(w1 a ) G →w c N 2 5 ^ a ,w & 1G GD ~ G w ! ad ~ G w ! ac
D →w d
~A8!
.
w : Another useful expression can be found by contracting relation ~A7! with G ab 21 * a ,w N 2 a ,w a →(w2 a ) A N 2 ~ G w ! ab ~ G w2 a ! AB
b → ~ w2 a ! B
21 w1 a * a , ~ w2 a ! N 2 a ,(w1 a ) G →w c N 2 5 ^ a ,w & D w 1G GD ~ G w ! cd
D →w d
~A9!
.
This expression is easily iterated along a sequence of lowerings with the same ladder operator: 21 * a ,w N 2 a ,w a →(w2 a ) A N 2 ~ G w ! ab ~ G w2 a ! AB
b → ~ w2 a ! B
21 a * a , ~ w12 a ! N 2 a ,(w12 a ) g →(w1 a ) G N 2 5 ^ a ,w & D w 1 ^ a ,w1 a & D w1 a 1G gw12 ~ G w1 a ! GD d
d → ~ w1 a ! D
5 ^ a ,w & D w 1 ^ a ,w1 a & D w1 a 1¯1 ^ a ,w1k a & D w1k a ,
~A10!
where (w1k a ) is the highest weight in the SU(2) subgroup chain (w), (w1 a ), (w12 a ),... present in the weight system. Finally, for completeness, when raising operators are applied, Eq. ~A8! can be written as 21 N a ,w a →(w1 a ) G N a*,w ~ G w1 a ! GD
a → ~ w1 a ! D
21 21 w2 a N a ,(w2 a ) C →w c N a*, ~ w2 a ! 52 ^ a ,w & 1G CD ~ G w ! ad ~ G w ! ac
D →w d
.
~A11!
Special Cases: Lowering within basis states. Consider a series of states connected by repeated application of the same lowering operator E 2 a . Choose any states with degenerate weights obtained in this series as part of the basis for the degenerate weights and label these states by i. For the sequence: ...,(w1 a ) i ,w i ,(w2 a ) i ,... the generalized recursion relation ~A8! reduces to w1 a w 21 * a , ~ w1 a ! u N 2 a ,w i →(w2 a ) i u 2 5 ^ a ,w & 1G GD ~ G w ! 21 ic ~ G ! id N 2 a ,(w1 a ) G →w c N 2
D →w d
. ~A12!
When (w1 a ) i is the only state which can be lowered by E 2 a to obtain a state of weight w we get a) u N 2 a ,w i →(w2 a ) i u 2 5 ^ a ,w & 1G (w1 u N 2 a ,(w1 a ) i →w i u 2 . ii
~A13!
This includes the case of a nondegenerate (w1 a ) weight subspace. When (w1 a ) is nondegena) 51, which further simplifies the above-given relation. erate G (w1 ii A special case of interest is lowering the degenerate zero weight states of the adjoint representation which correspond to the Cartan subalgebra. These degenerate weight states can be labeled u (0) i & where the ith degenerate weight is obtained by E 2 a i u a i & } u (0) i & . This basis, however, is not orthogonal. When lowering such a basis state the general formula ~A8! reduces to 2 2 2 2 u N 2 a i ,(0) j →(2 a i ) u 2 5 @~ G (0) ! 21 ji # u N 2 a i ,( a i )→(0) i u 5 ^ ~ 0 ! j u ~ 0 ! i & u N 2 a i ,( a i )→(0) i u .
~A14!
This result is consistent with formula ~3! in Sec. II when applied to the zero weight states of the 78 in E 6 .
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp
013506-13
E 6 unification model building. III
J. Chem. Phys. 46, 013506 (2005)
1
F. Gursey, P. Ramond, and P. Sikivie, Phys. Lett. 60B, 177 ~1976!. C. D. Froggatt and H. B. Nielsen, Nucl. Phys. B 147, 277 ~1979!. 3 S.-P. Li, R. V. Moody, M. Nicolescu, and J. Patera, J. Math. Phys. 27, 668 ~1986!. 4 I.-G. Koh, J. Patera, and C. Rousseau, J. Math. Phys. 25, 2863 ~1984!. 5 G. Anderson and T. Blazˇek, J. Math. Phys. 41, 4808 ~2000!. 6 G. Anderson and T. Blazˇek, J. Math. Phys. 41, 8170 ~2000!. 7 M. Bando, T. Kugo, and K. Yoshioka, Prog. Theor. Phys. 104, 211 ~2000!. 8 R. Slansky, Phys. Rep. 79, 1 ~1981!. 9 For roots and weights we use the notation where ¯x [2x. 10 R. V. Moody and J. Patera, J. Math. Phys. 25, 2838 ~1984!. 11 Some applications of the charge conjugation operators can be found in Ref. 5. 12 G. Anderson and T. Blazˇek ~unpublished!. 13 It is not always possible to choose a single basis for all of the states of degenerate weights in a representation such that lowering a particular state results in a particular state of lower weight ~as opposed to a linear combination of weights! for all possible lowerings. The simplest example of this is the 27 dimensional representation of SU(3). 14 See EPAPS Document No. E-JMAPAQ-44-021203 for all 31 tables. A direct link to this document may be found in the online article’s HTML reference section. This document may also be reached via the EPAPS home page ~http:// www.aip.org/pubservs/epaps.html! or from ftp.aip.org in the directory /epaps/. See the EPAPS homepage for more information. 2
Downloaded 06 Aug 2008 to 195.128.96.210. Redistribution subject to AIP license or copyright; see http://jmp.aip.org/jmp/copyright.jsp